Compuware has undergone a transformation in recent months, some of which has been covered in my colleague Bola Rotibi’s January 2014 blog “Out with the old and in with the new: Compuware 2.0 no more”. . With the divestiture of the Uniface, Changepoint and Professional services business units to Marlin Equity Partners for $160 million, Compuware is left with its mainframe product and services and the application performance management solutions.
Once everything shakes out the company will look very different to what it did a year or so ago. While they will retain two product categories the future of the business surely revolves around Compuware APM. Application Performance Management (APM) is a suite of tools that provide development organisations with the ability to monitor their applications in production for performance problems. These might include slow running code, poor performing database queries or network delays.
Such performance monitoring has been very fashionable of late. From developer tools to web browsers there has been a concerted drive to provide developers with detailed metrics about how their application is running. The recent mobile revolution has fuelled this as mobile is far more unforgiving when it comes to performance than traditional software experiences. A user might be okay to sit at their desktop and wait a few seconds (or longer) for a task to complete but not when using a mobile app.
The challenge for many of these new performance monitoring capabilities has been the complexity of modern applications. It is great that some of the latest web browsers can provide performance profiles but they only cover what’s happening in the browser. What the application is doing on the server is outside of their realm. Therefore developers are only getting part of the complete picture when it comes to application performance. Other tools have similar issues as they take their piece of the application and not the whole.
This is where Compuware APM finds its strength. The product can monitor performance from the user interface (whatever that might be: desktop, browser, mobile and so on) to the database. Developers are provided with a complete picture of what their application is doing and where any problems (or potential problems) might be.
A further challenge for Compuware is that because the product touches the entire application from front to back, so who is its customer?
Of course, Compuware is not alone in this market with products from bigger vendors such as IBM and HP and also smaller start-ups like New Relic. However, the important question to ask is how do they then position their products and services against the growing noise around operational intelligence with many players – large and small, established and new entrants – laying claim to deliver in part some of that application performance management capability.
The smaller guys may actually prove to be the bigger challenge as they are proving increasingly popular with smaller but also some larger development shops. They certainly seemed to have achieved levels of awareness and market penetration faster than many of the established vendors, including Compuware, who have been in this space a lot longer.
Symptomatic of a larger problem
This raises a bigger question about software development and delivery today. In many enterprises there is a separation of roles. DBAs will be responsible for the database, developers the middleware and frontend code, and IT operations for the infrastructure, deployment and management of the application. They each have their own focus. A DBA will be concerned about query times whereas someone in IT operations will focus on how the hardware is performing. The result is often that when applications perform badly somewhere, knowledge of it falls between the cracks of ownership and diagnosis proves difficult because there is not a holistic picture to look at.
Hence we end up with the aforementioned tools that serve particular audiences. Compuware is therefore selling to multiple audiences when essentially they need to sell to the entire development organisation. How to do this is tricky and they have certainly identified some particular use cases such as DevOps. DevOps brings together development and operations and so provides a nice fit for their product. Unfortunately we are a long way off DevOps being mainstream and so Compuware find themselves far ahead of the market in this area. They are also not alone in jumping on the DevOps band wagon.
The more interesting pitch is around User Experience (UX). UX is widely misunderstood within businesses, especially enterprises, as being a design discipline. If you want your application to look “pretty” then you apply UX. This is far from the truth and Compuware have recognised this. An application’s experience covers not just the visual but also usability, process, security, reliability, scalability and performance. The best looking app in the world will be rejected by users if it is interminably slow to do anything or opens them up to security issues
Hitching the APM wagon to UX
Driven largely by mobile which demands better experiences for users, organisations are beginning to recognise the importance of UX and all of its constituent parts. The highly unusable, slow and buggy corporate ERP system is no longer acceptable to not just its users but the business itself. They are beginning to understand how it negatively impacts the efficiency and effectiveness of the business as well as impacting business performance, competitiveness, customer experience and other key factors that inevitably hit the bottom line.
The result is that the business demands better from IT. To deliver that, IT must change the way that it builds, deploys and manages applications. All parts of the process need to be far more integrated and the touch points between much more frictionless. We saw this in our recent case study of Barclay’s PingIt mobile application. To enable this they need a single pain of glass into the application and its performance. It is no good just to be concerned about one aspect but to see why an application performs the way that it does across the stack.
A big challenge when I was working with organisations to improve UX was finding the tools that would provide this holistic view. Compuware APM certainly does a lot to address the need and in a world where applications might be installed on the desktop, running in a web browser or on a mobile device.
To be a true UX tool there is work to be done, mainly around usability metrics and perhaps even user testing. There are products that address these specific needs, but Compuware has built the harder part. Following in the wake of UX is therefore a good way for Compuware to find new customers. But similar to DevOps they will find themselves ahead of the customer. The challenge for them will be whether they wait for the customer to catch-up, or whether they step up to promote and evangelise UX in order to advance its march.
No room for historical mistakes: Marketing with the correct terminology is everything
What’s odd is that Compuware APM, and tools just like it, should be a no-brainer to any development organisation. Perhaps this is why newer vendors are having success in newer development environments such as start-ups who see the value. Having spent most of the last 20 years in development I can attest that such a tool would have proven its worth time and again. The fact that everyone does not buy into these tools is testament to a quality problem within the industry.
Compuware has to accept some responsibility as well. It is not like I’ve not looked for such tools in the past. However, more common terms are “performance monitoring” or “performance testing” rather than APM. Put either of the former terms into Google and Compuware does not come up for a few pages which is the same as not coming up at all. IBM, Microsoft, HP and more recent players such as New Relic all do better. As well as chasing markets they would do well to address some messaging basics that would help those actually looking for their product to find them.